RECONSTRUCTIVE

A Prospective Cohort Study of Fibula Free Flap Donor-Site Morbidity in 157 Consecutive Patients
Adeyiza O. Momoh, M.D. Peirong Yu, M.D. Roman J. Skoracki, M.D. Suyu Liu, M.S. Lei Feng, M.S. Matthew M. Hanasono, M.D.
Houston, Texas

Background: Although the fibula free flap is preferred for bony head and neck reconstruction, donor-site morbidity remains a concern. The authors goal was to evaluate potential risk factors for complications and whether the type of wound closure and timing of postoperative ambulation had an effect on the development of short- and long-term morbidities. Methods: A prospective cohort study of donor-site morbidity was performed in 157 consecutive patients who underwent fibula free flap reconstruction for head and neck defects. Results: Perioperative donor-site complications occurred in 31.2 percent of patients, including skin graft loss (15 percent), cellulitis (10 percent), wound dehiscence (8 percent), and abscess (1 percent). Preoperative chemotherapy (p 0.02) was associated with increased complications. No significant difference in complication rates was observed between primary and skin graft wound closure (p 0.59). The timing of ambulation was not related to the development of complications (p 0.41). Long-term morbidities occurred in 17 percent of patients and included leg weakness (8 percent), ankle instability (4 percent), great toe contracture (9 percent), and decreased ankle mobility (12 percent). The occurrence of perioperative complications, flap type, and closure technique were not significantly associated with long-term morbidities. Functionally, 96 percent of patients returned to their preoperative level of ambulatory activity. Decreases in ambulatory status could all be ascribed to causes other than donor-site morbidity. Conclusion: Fibula free flap harvest is associated with a high rate of complications, but the majority of patients have no long-term functional limitations. (Plast. Reconstr. Surg. 128: 714, 2011.) CLINICAL QUESTION/LEVEL OF EVIDENCE: Risk, IV.

lthough the fibula free flap is the preferred flap for bony head and neck reconstruction in most centers, donor-site morbidity remains a concern.1,2 The perioperative donor-site complication rate associated with the fibula free flap has been reported to be as little as 2 percent and as much as 38 percent.37 Reported long-term donor-site morbidity has also been variable, with some studies reporting no long-term morbidity and others suggesting that the majority of patients experience long-term problems with joint stiffness and instability, muscular weakness, or gait
From the Departments of Plastic Surgery and Biostatistics, University of Texas M. D. Anderson Cancer Center. Received for publication December 2, 2010; accepted March 18, 2011. Presented at the 89th Annual Meeting of the American Association of Plastic Surgeons, in San Antonio, Texas, March 20 through 23, 2010. Copyright 2011 by the American Society of Plastic Surgeons DOI: 10.1097/PRS.0b013e318221dc2a

abnormalities.4,6 13 Given the prominent role of the fibula free flap in contemporary head and neck reconstruction, we felt that it was important to prospectively identify short- and long-term morbidities associated with this flap. In addition, we sought to identify predictive factors for complications and poor functional outcomes and to evaluate various options in how the donor site is managed operatively and postoperatively, which have rarely been studied before. Subjects of debate we address include the effect of method of donor-site closure (i.e., primary versus skin graft closure) and

Disclosure: The authors have no commercial associations or financial disclosures that might pose or create a conflict of interest with information presented in this article. No funding was received for the work presented in this article.

714

www.PRSJournal.com

Volume 128, Number 3 Free Fibula Flap Donor Site

the timing of weight-bearing ambulation following surgery. formed through the soleus and flexor hallucis longus muscles, if necessary, rather than removing some or all of these muscles to protect the perforators. All donor-site wounds were closed either primarily or with a split-thickness skin graft.18 The decision to close the donor wound primarily or to perform skin graft closure was made based on whether there was more than a moderate amount of tension on the wound in doing so. This was admittedly a subjective decision, as is routinely the case in most wound closures. Skin grafts were covered with a bolster dressing and left in place for 5 days. After bolster removal, antibiotic ointment was applied to the graft, which was then covered with nonadherent petroleum gauze, followed by a gauze dressing, and then wrapped in an elastic compression bandage until fully healed. After complete healing of the incision and skin graft (if performed), compression stockings were recommended until lower leg edema resolved. A plaster posterior splint was applied in all patients, with the ankle in 90 degrees of dorsiflexion, and removed on postoperative day 5, at the same time as bolster removal when skin graft closure was performed. While in bed, the leg that had been operated on was elevated continuously. Patients were allowed to sit in a chair with the donor leg elevated as early as 24 hours after surgery. The timing of ambulation was examined as a possible contributor to the development of perioperative complications. Patients were instructed to ambulate at different times based on surgeon preference. However, other factors also played a role (e.g., whether the patient was unable or unwilling to ambulate for various reasons or whether physical therapy was available or not). Therefore, the actual day when ambulation began was recorded. When ambulation was begun, weight bearing was allowed as tolerated. The typical time for ambulation was 15 minutes to start, advancing the total ambulation time as tolerated by the patient. Patients were instructed to elevate their donor-site leg whenever not ambulating, to minimize edema. Perioperative complications, occurring within 30 days after surgery, were analyzed separately from long-term ( 30 days) morbidities. The Wilcoxon rank sum test was used to compare continuous variables, because they were not normally distributed. The chi-square test or Fishers exact test was used as appropriate for categorical variable comparisons. All tests were two-sided, and values of p 0.05 were considered significant. Values of p 0.0001 are stated as such; otherwise, exact p values are given.

PATIENTS AND METHODS

The study cohort was composed of consecutive patients undergoing fibula free flap harvest performed by the authors (M.M.H., R.J.S., and P.Y.) at the University of Texas M. D. Anderson Cancer Center between 2005 and 2009. Data specifically regarding donor-site morbidity were collected prospectively by the authors during the immediate postoperative period and at each follow-up appointment, and maintained in a database specific to this study. Patients with less than 3 months of follow-up for various reasons (e.g., death, tumor recurrence, hardship precluding travel) were excluded from the study. Institutional review board approval was obtained before this study was undertaken. A preoperative history and physical examination was performed for each patient, and information regarding comorbidities and other potential risk factors for complications was recorded, including advanced age, tobacco use, and preoperative chemotherapy. The need for angiography was also noted. Conventional or magnetic resonance lower extremity angiograms were obtained in patients with decreased pedal pulses or with symptoms or signs of vascular insufficiency, such as claudication, cool lower limb, and atrophic or mottled skin, to verify patent distal circulation as recommended by others.14 16 A prior diagnosis of peripheral vascular disease by another physician was also an indication to obtain an angiogram. Information regarding donor-site complications occurring during the perioperative period ( 30 days after surgery) was recorded. Lower extremity subjective complaints (i.e., muscular weakness and/or joint instability), objective abnormalities (i.e., decreased range of motion and/or joint contractures), and ambulatory status (i.e., nonambulatory, ambulatory with the assistance of a walker or cane, ambulatory without assistance but with an abnormal gait, ambulatory without assistance, and able to run or participate in athletics) were assessed before surgery and at each follow-up visit. Patients were seen 1 week after discharge, 1 month after discharge, and at approximately 3-month intervals for at least 1 year. Free fibula flap dissection was performed as described previously by others.17,18 Five centimeters of proximal bone and 7 cm of distal bone were left in situ to preserve knee and ankle stability, respectively. When an osteocutaneous free flap was harvested, a perforator dissection was per-

715

Plastic and Reconstructive Surgery September 2011 RESULTS

During the study period, 165 consecutive fibula free flaps were performed for head and neck reconstructions. Eight patients with less than 3 months of follow-up were excluded from the study. Reasons for lack of follow-up included five deaths caused by rapid recurrence of disease and three deaths resulting from other causes. The remaining 157 patients included 103 male patients and 54 female patients, with a mean age of 53 17 years (range, 5 to 87 years). The mean length of follow-up in this group of patients was 24.9 15.2 months (range, 3 to 54 months) after surgery. There were no flap losses in this series. Osseous flaps were harvested in 19 patients (12 percent) and osteocutaneous flaps were harvested in the remaining 138 patients (88 percent). Fiftysix donor sites (36 percent) were closed primarily and 101 (64 percent) were closed using a skin graft. All osseous free flap donor sites were closed primarily. The mean width of osteocutaneous free flap donor sites closed primarily was 3.6 0.74 cm (range, 2 to 5 cm). The mean width of osteocutaneous free flap donor sites closed with a skin graft was 5.3 1.5 cm (range, 3 to 10 cm), which was significantly different from the width that was closed primarily (p 0.0001). Donor-site complications are listed in Table 1. Of the 13 patients with wound dehiscences, three (23 percent) developed tendon exposure. Of the 24 patients with partial or total skin graft loss, eight (33 percent) developed tendon exposure. Five patients (3 percent) required operative intervention for their complication. The remainder healed with antibiotics and/or local wound care, including all but three patients with tendon exposure. The average time to healing after dehiscence or skin graft loss was 11.8 7.6 weeks (range, 3 to 32 weeks). The association of various medical comorbidities and other potential risk factors with the occurrence of perioperative donor-site complications is shown in Table 2. There was an increased incidence
Table 1. Perioperative Donor-Site Complications in 157 Patients Undergoing Fibula Free Flap Reconstruction
Complications Partial graft loss Cellulitis Wound dehiscence Total graft loss Abscess Total No. of Patients (%) 22 (14) 15 (10) 13 (8) 2 (1) 1 (1) 49 (31)*

Table 2. Association of Donor-Site Complications with Medical Comorbidities and Other Potential Risk Factors in Patients Undergoing Fibula Free Flap Reconstruction (n 157)
No. of Patients (%) 15 (10) 12 (8) 18 (11) 6 (4) 3 (2) 19 (12) 46 (29) 39 (25) 16 (10) No. of Patients with Complication (%) 4 (27) 3 (25) 7 (39) 1 (17) 2 (67) 9 (47) 15 (32) 6 (15) 4 (25)

Comorbidity/ Risk Factor Cardiac disease Pulmonary disease Diabetes mellitus Peripheral vascular disease Morbid obesity Age 70 years Current tobacco use Preoperative chemotherapy Angiography*

p 0.43 0.75 0.45 0.67 0.23 0.10 0.84 0.02 0.31

*Patients for whom angiography was ordered because of a history suggesting vascular insufficiency or an abnormal lower extremity physical examination.

of complications in patients who received preoperative chemotherapy (p 0.02). Each complication was also analyzed individually to determine whether any comorbidity or risk factor was associated with the occurrence of that complication (data not shown). The need for preoperative angiography, based on a history suggesting vascular insufficiency or an abnormal physical examination, was found to be associated with wound dehiscence (p 0.03), and the association with a prior diagnosis of peripheral vascular disease approached but did not reach statistical significance (p 0.05). The comorbidities and risk factors examined were not associated with the occurrence of any other complications. The association between the occurrence of complications and the type of flap and the management of the donor site is shown in Table 3. An analysis of complications occurring relative to the area of the donor-site wound that needed to be skin grafted showed that larger skin-grafted donor wounds ( 100 cm2) were more likely to exhibit skin graft loss
Table 3. Association of Donor-Site Complications with Flap Type and Donor-Site Management in Patients Undergoing Fibula Free Flap Reconstruction (n 157)
Flap Type/ Donor-Site Management Flap type Osseous Osteocutaneous Type of closure Primary Skin graft No. of Patients (%) 19 (12) 138 (88) 56 (36) 101 (64) No. of Patients with Complication (%) 2 (11) 47 (35) 0.60 16 (29) 33 (33)

p 0.06

*Four patients experienced more than one perioperative complication.

716

Volume 128, Number 3 Free Fibula Flap Donor Site

(p 0.04) and tendon exposure (p 0.007) than smaller donor wounds ( 100 cm2). However, complications were not more likely to occur with either primary or skin graft wound closure (p 0.60). Even when only looking at osteocutaneous free flaps, the rate of donor-site complications for primary closure (36 percent) was not significantly different from the rate of donor-site complications for skin graft closure (32 percent) (p 0.69). Patients who developed a complication began ambulating at a mean of 5.6 2.2 days after surgery, whereas patients who did not develop a complication began ambulating at a mean of 5.2 2.1 days after surgery (p 0.41). The type of flap, type of wound closure, and the timing of ambulation were not related to the occurrence of any one complication when analyzed as a risk factor for each type of complication individually (data not shown). Subjective and objective long-term morbidities are listed in Table 4. No patients experienced knee instability, weakness, or decreased range of motion. The occurrence of perioperative complications was not associated with the development of long-term morbidities (p 0.90). The type of flap (osseous or osteocutaneous) and type of closure (primary or skin graft) were also not significantly related to the development of long-term morbidities (p 0.19 and p 0.68, respectively). Preoperative and postoperative ambulatory statuses are compared in Table 5. A decrease in the
Table 4. Long-Term Morbidity in Patients Undergoing Fibula Free Flap Reconstruction (n 157)
Morbidity Leg weakness Ankle instability Great toe flexion contracture Decreased ankle range of motion Total No. of Patients (%)* 12 (8) 6 (4) 14 (9) 3 (2) 26 (17)

level of ambulatory function was observed in seven patients (4 percent). However, all decreases in ambulatory status were attributable to causes other than donor-site morbidity: five patients experienced a decrease in ambulatory status caused by debilitation associated with treatment for recurrent disease, one patient underwent a partial lung resection for a second primary lung cancer limiting pulmonary function, and one patient had trauma to the knee requiring a knee replacement. Table 6 shows the number of patients who returned to their preoperative ambulatory status as a function of the number of months after surgery, based on their status at the time of their follow-up appointments at 1, 3, 6, 9, and 12 months. Besides the seven patients who experienced a decline in ambulatory status, 18 patients were excluded from this analysis because they missed one or more follow-up appointments. By 12 months after surgery, all but the seven patients who experienced a decline in ambulatory status had returned to their preoperative level of activity.

DISCUSSION
The overall rate of perioperative complications following fibula free flap harvest was 31 percent in the present study. Only 3 percent of the complications required operative intervention to resolve. However, healing for some complications took a prolonged period, up to 32 weeks. Our perioperative complication rate was consistent with the complication rates reported in some studies but not others, which reported much lower rates. Studies of fibula free flap donor-site morbidity following head and neck reconstruction are summarized in Table 7. In contrast to the present study, prior studies have been retrospective in nature and limited to smaller numbers of patients. Similar to the present study, the most common perioperative comTable 6. Return to Preoperative Ambulatory Status as a Function of Time in Patients Undergoing Fibula Free Flap Reconstruction (n 132)*
Months after Surgery 1 3 6 9 12 No. of Patients Who Returned to Preoperative Status (%) 32 (24) 53 (40) 33 (25) 11 (8) 2 (2) Cumulative No. of Patients Who Returned to Preoperative Status (%) 32 (24) 85 (64) 118 (89) 129 (98) 132 (100)

*Seven patients experienced more than one long-term morbidity.

Table 5. Comparison of Preoperative and Postoperative Ambulatory Status in Patients Undergoing Fibula Free Flap Reconstruction (n
Ambulatory Status Nonambulatory Ambulatory with cane or walker Ambulatory without assistance but abnormal gait Normal ambulation Running or athletics Preoperative (%) 0 (0) 3 (2) 0 (0) 93 (59) 61 (39)

157)

Postoperative (%)* 0 (0) 6 (4) 1 (1) 92 (59) 58 (37)

*Seven patients experienced a decline in ambulatory status; however, all of these changes could be ascribed to causes other than donor-site morbidity.

*Seven patients who experienced a decline in ambulatory status for recurrent cancer, lung surgery, or knee surgery were excluded from this analysis. An additional 18 patients who missed one or more follow-up appointments were also excluded.

717

Plastic and Reconstructive Surgery September 2011

Table 7. Comparison of Studies Involving Donor-Site Morbidity following Fibula Free Flap Harvest
Reference Anthony et al., 19953 Shpitzer et al., 19974 Shindo et al., 20005 Zimmerman et al., 20017 Momoh et al.* Study Type Retrospective Retrospective Retrospective Retrospective Prospective No. of Patients 30 41 55 42 157 Length of Follow-Up (mo) 7.3 17 NA 35 18.4 Perioperative Complications (%) 17 2 28 38 31 Long-Term Morbidity (%) NA 27 NA 25 17

NA, not available or cannot be determined from the data presented. *Present study.

plications include partial or total skin graft loss, wound dehiscence, and infection. There have been few prior analyses of risk factors for donor-site morbidity associated with the fibula free flap. In the present study, preoperative chemotherapy was found to be a significant predisposing factor for perioperative complications. Peripheral vascular disease may also be a risk factor for wound dehiscence but not statistically significantly so. Notably, two of six patients with a prior diagnosis of peripheral vascular disease had unremarkable lower extremity physical examinations, and four of six patients with a prior diagnosis of peripheral vascular disease underwent angiography that showed patent three-vessel flow to the distal leg. One reason that the diagnosis of peripheral vascular disease and other conditions associated with compromised tissue vascularity, such as diabetes and smoking, may not have been found to be strongly associated with complications is that symptoms and signs of advanced vascular insufficiency may have eliminated patients from undergoing fibula free flap reconstruction entirely. Consistent with this is that the need for angiography, based on abnormal findings during the history and physical examination, was itself found to be predictive of donor-site wound dehiscence. Looked at another way, the findings of this study suggest that patients without signs or symptoms of peripheral vascular compromise, either by prior diagnosis or by preoperative evaluation, can undergo fibula free flap harvest without elevated risks for donor-site morbidity (with the exception of those who have undergone neoadjuvant chemotherapy). Some authors have suggested that skin grafting may result in higher complication rates because of poor graft take of the donor-site wound bed.4,8 Others have hypothesized that moderately tight primary closures result in increased tissue ischemia and injury and therefore advocate skin graft closure of all but the smallest of donor-site wounds to avoid wound dehiscence or late morbidities related to tissue fibrosis.5 Only one prior

study has specifically examined the relationship of wound closure technique to the occurrence of donor-site morbidity: Shindo et al.5 found an 18 percent complication rate in donor sites closed with a skin graft compared with a 38 percent complication rate in those closed primarily, and ascribed the difference to the possible development of a pseudocompartment syndrome in donor sites closed primarily, although the difference in complication rates was not significant (p 0.10). Overall, 33 percent of patients in the present study who had their donor sites skin grafted experienced perioperative complications compared with 29 percent of patients who had primary wound closure (p 0.59). Our data therefore do not support prophylactically skin grafting donor-site wounds that can be closed under modest tension, using proper surgical judgment and common sense. The timing of ambulation following fibula harvest has never previously been studied. The few published protocols for ambulation following fibula free flap harvest vary considerably and are not evidence-based. For example, Zimmermann et al.7 allow crutch walking on postoperative day 1 and allow the patient to progress as tolerated to unassisted full-weight-bearing ambulation. Babovic et al.6 allow ambulation on postoperative day 2 in patients who undergo primary closure, but require bed rest until postoperative day 5 in patients who undergo skin graft closure, at which time the skin graft dressings are removed and weight-bearing ambulation is allowed. Papadopulos et al.19 mobilize their patients on the fifth postoperative day, allowing weight bearing as tolerated, although a splint is kept on for 6 weeks after surgery. Because we did not observe a significant association between the incidence of complications and timing of first ambulation, we now encourage early mobilization of our patients, beginning on postoperative day 1 or 2, if possible, because this may theoretically minimize the risks associated with prolonged bed rest, such as atelectasis, pneumonia, venous thromboembolism, and pressure ulcers.

718

Volume 128, Number 3 Free Fibula Flap Donor Site

In this study, a few patients complained of leg weakness (8 percent) and ankle instability (4 percent). Physical examination revealed some limitations in ankle range of motion (2 percent) and great toe range of motion (9 percent). Comparison with prior reports (Table 7) is difficult because investigators may inquire about different symptoms and assess function differently; however, the morbidities we elicited are consistent with those found by others.35,7 Leg weakness and decreased ankle range of motion are most likely caused by muscle damage and postsurgical fibrosis. Great toe stiffness or clawing is attributed to detachment of the flexor hallucis longus muscle and subsequent scarring.13 Some improvement in our series over others may reflect our practice of performing perforator dissection, in theory preserving the vascularity of the flexor hallucis longus muscle, rather than removing a significant portion of soleus and flexor hallucis longus muscles to maintain skin paddle blood supply, which was a more common practice previously.19,20 This study may lack the sensitivity to detect subtle physical limitations associated with fibula free flap harvest. A few smaller sized studies have examined range of motion and strength following fibula free flap harvest more intensively and have reported measurable differences between the legs that had and had not been operated on. For example, in a study involving 11 patients who underwent isokinetic testing following surgery, Anthony et al.3 found a significant difference in ankle flexion and extension range of motion (p 0.0033) and significant differences in knee flexion and extension strength (p 0.0264 and p 0.0019, respectively), and ankle plantar flexion and dorsiflexion strength (p 0.0085 and p 0.0174, respectively) compared with the leg that had not been operated on. Rogers et al.11 similarly found a 10- to 25-degree decrease in range of ankle dorsiflexion compared with the normal side in five of 11 patients tested. In contrast, Tang et al.10 failed to detect any significant differences in knee or ankle range of motion in 39 fibula free flap patients undergoing similar testing. From a functional standpoint, most studies, including the present one, have shown that all or nearly all patients are able to successfully perform their activities of daily living without significant limitations.4,8,12 In fact, 37 percent of the patients in this study reported being able to run or engage in athletic activity following surgery. Even during testing with specialized equipment, functional limitations are generally difficult to elicit during normal ambulation. For example, although Chou et al.13 found differences in complex tests of balance and gait, such as maintaining balance with the eyes closed and swaying support, no significant differences in balance or gait between 11 subjects who had undergone fibula free flap harvest and 10 control subjects were noted under straightforward conditions. Thus, although mild deficits probably occur with some frequency, useful function is always or almost always preserved.

CONCLUSIONS
Fibula free flap harvest is associated with a high rate of complications. However, complications requiring surgical intervention are rare, and the vast majority of patients have no long-term functional limitations. Neither primary closure nor skin grafting is associated with an increased level of complications. Early weight-bearing ambulation is encouraged following fibula surgery because an association between the development of complications and the timing of postoperative ambulation was not observed.
Matthew M. Hanasono, M.D. Department of Plastic Surgery, Unit 443 University of Texas M. D. Anderson Cancer Center 1515 Holcombe Boulevard Houston, Texas 77030 mhanasono@mdanderson.org

REFERENCES
1. Wallace CG, Chang YM, Tsai CY, Wei FC. Harnessing the potential of the free fibula osteoseptocutaneous flap in mandible reconstruction. Plast Reconstr Surg. 2010;125:305313. 2. Cordeiro PG, Disa JJ, Hidalgo DA, Hu QY. Reconstruction of the mandible with osseous free flaps: A 10-year experience with 150 consecutive patients. Plast Reconstr Surg. 1999;104: 13141320. 3. Anthony JP, Rawnsley JD, Benhaim P, Ritter EF, Sadowsky SH, Singer MI. Donor leg morbidity and function after fibula free flap mandible reconstruction. Plast Reconstr Surg. 1995; 96:146152. 4. Shpitzer T, Neligan P, Boyd B, Gullane P, Gur E, Freeman J. Leg morbidity and function following fibula free flap harvest. Ann Plast Surg. 1997;38:460464. 5. Shindo M, Fong BP, Funk GF, Karnell LH. The fibula osteocutaneous flap in head and neck reconstruction: A critical evaluation of donor site morbidity. Arch Otolaryngol Head Neck Surg. 2000;126:14671472. 6. Babovic S, Johnson CH, Finical SJ. Free fibula donor-site morbidity: The Mayo experience with 100 consecutive harvests. J Reconstr Microsurg. 2000;16:107110. 7. Zimmermann CE, Borner BI, Hasse A, Sieg P. Donor site morbidity after microvascular fibula transfer. Clin Oral Investig. 2001;5:214219. 8. Hidalgo DA, Rekow A. A review of 60 consecutive fibula free flap mandible reconstructions. Plast Reconstr Surg. 1995;96: 585596; discussion 597602. 9. Vail TP, Urbaniak JR. Donor-site morbidity with use of vascularized autogenous fibular grafts. J Bone Joint Surg Am. 1996;78:204211.

719

Plastic and Reconstructive Surgery September 2011

10. Tang CL, Mahoney J, McKee MD, Richards RR, Waddell JP, Louie B. Donor site morbidity following vascularized fibular grafting. Microsurgery 1998;18:383386. 11. Rogers SN, Lakshmiah SR, Narayan B, et al. A comparison of the long-term morbidity following deep circumflex iliac and fibula free flaps for reconstruction following head and neck cancer. Plast Reconstr Surg. 2003;112:15171525; discussion 15261527. 12. Bodde EW, de Visser E, Duysens JE, Hartman EH. Donor-site morbidity after free vascularized autogenous fibular transfer: Subjective and quantitative analyses. Plast Reconstr Surg. 2003; 111:22372242. 13. Chou SW, Liao HT, Yazar S, Lin CH, Lin YC, Wei FC. Assessment of fibula osteoseptocutaneous flap donor-site morbidity using balance and gait test. J Orthop Res. 2009;27:555560. 14. Disa JJ, Cordeiro PG. The current role of preoperative arteriography in free fibula flaps. Plast Reconstr Surg. 1998;102: 10831088. 15. Lutz BS, Wei FC, Ng SH, Chen IH, Chen SH. Routine donor leg angiography before vascularized free fibula transplantation is not necessary: A prospective study in 120 clinical cases. Plast Reconstr Surg. 1999;103:121127. Dublin BA, Karp NS, Kasabian AK, Kolker AR, Shah MH. Selective use of preoperative lower extremity arteriography in free flap reconstruction. Ann Plast Surg. 1997;38:404407. Wei FC, Seah CS, Tsai YC, Liu SJ, Tsai MS. Fibula osteoseptocutaneous flap for reconstruction of composite mandibular defects. Plast Reconstr Surg. 1994;93:294304; discussion 305306. Jones NF, Monstrey S, Gambier BA. Reliability of the fibular osteocutaneous flap for mandibular reconstruction: Anatomical and surgical confirmation. Plast Reconstr Surg. 1996; 97:707716; discussion 717718. Papadopulos NA, Schaff J, Bucher H, Groener R, Geishauser M, Biemer E. Donor site morbidity after harvest of free osteofasciocutaneous fibular flaps with an extended skin island. Ann Plast Surg. 2002;49:138144. Schusterman MA, Reece GP, Miller MJ, Harris S. The osteocutaneous free fibula flap: Is the skin paddle reliable? Plast Reconstr Surg. 1992;90:787793; discussion 794798.

16.

17.

18.

19.

20.

720